Skip to contents

Population change map

Source: vignettes/a01_Interactive-map.Rmd
a01_Interactive-map.Rmd

Use the interactive map below to check population changes. Zoom in to see where Podocnemis unifilis is Endangered based on IUCN Red List criteria - A3bd.

  • Overall Podocnemis unifilis is Endangered (“Em perigo” / “En peligro”) based on future population size reduction criteria - A3bd.
    Within 3 generations (35 years) the adult female population is predicted to decline by 50.6% in the future (25 and 75% quantile range: 64.7 - 34.4 % decline).

Map

If you find any errors (e.g. points where Podocnemis unifilis does not occur, points where species is “Endangered” but populations are increasing etc) please raise an issue at: https://github.com/darrennorris/TACAR/issues .

  • When you zoom in you will see individual points.
    The points are brown where populations are predicted to decline by 50% or more within 3 generations (35 years). Brown points therefore represent rivers where the species is Endangered (“Em perigo” / “En peligro”), following the IUCN Red List population size reduction criteria - A3bd (https://www.iucnredlist.org/about/faqs). The values presented are from the “Business-as-Usual” (BAU) scenario. For a comparison among pessimistic, BAU and optimistic scenarios see the Basin-Country comparison figure below.

  • The points follow rivers mapped by remote sensing. This standardized global scale mapping comes from Grill et al 2019, Free-flowing Rivers: https://doi.org/10.1038/s41586-019-1111-9. The points are locations along rivers selected to represent where Podocnemis unifilis females are likely to nest and that are likely to be accessible to people by boat. Rivers were chosen within the species extent of occurance that had average long-term discharge rates equal to or greater than 15 cubic meters per second. This value corresponds to rivers that are accessible with nesting areas in a Brazilian study area with relatively small rivers (Bárcenas‐García et. al. 2021). To facilitate online viewing the mapped points are a subset at intervals of approximately 10 kilometers.

  • Due to the number of points, the map can become slow to respond when you zoom in.
    If this happens, zoom out to a level showing fewer points and you can pan around the map to find the area of interest. Then zoom in again to check the individual points. Lower zoom levels show locations of grouped points, with numbers and shaded coloring showing the number of individual point locations in each group.

  • The population reductions are conservative (best-case) estimates.
    Endangered (A3bd), is threfore a precautionary IUCN Red List assessment.

    • Important threats are not included (e.g. land use change Zalles et. al. 2021).
    • Losses are potentially buffered by an overly generous 20% ceiling to increases in unaccesible/unhunted populations.
    • Population projections use the earliest likely breeding age. The age at first reproduction was set to 5 years (Norris et. al. 2019).
    • Hunting of adult females was limited to 10% per year in the most pessimistic scenarios.
    • Hunting accessibility was limited to a distance of 48 km from locations with a human population density of 5 or more people per km2. This will underestimate impacts associated with threats such as large scale harvest around urban areas (Chaves et. al. 2021, Tregidgo et. al. 2017) and harvest in less densly populated rural areas (da Silva et al. 2022, Peres 2000).

Basin-country comparison

Here a graphical comparison of the projected population changes among basin-country combinations helps to provide a more detailed assessment.

In the above figure a population change of “-0.5” is 50% loss. The shaded shapes show predicted population changes for each basin-country combination. The labels to the right show river lengths in thousands of kilometers (e.g. “22.2 K km” is 22,200 kilometers). The vertical dashed line shows the overall mean value, and the verticle grey shaded area is the interquartile range (IQR - 25% and 75% quantile values). As the basin-country combinations cover different sized areas, estimates of overall mean and IQR are obtained using calculations weighted by the river length.

Within each basin-country combination, the horizontal black crossbars show the median and interquartile range of projection scenarios. These scenarios represent the impact of threats (e.g. hunting) across three generations. The impact of threats ranges from most severe (“pessimistic”), Business-as-usual (“BAU”), to least severe (“optimistic”). Generation length was estimated from stage-based matrix population models (Bienvenu & Legendre, 2015). The scenario results include the mean (41 year), and the interquartile range of three generations (35 and 45 year, 25% and 75% quantile values respectively). Generation length was estimated using the Matrix Projection Models (Bienvenu & Legendre 2015). The estimates follow the IUCN definition: “Generation length is the average age of parents of the current cohort (i.e., newborn individuals in the population). Generation length therefore reflects the turnover rate of breeding individuals in a population….. Where generation length varies under threat, such as the exploitation of fishes, the more natural, i.e. pre-disturbance, generation length should be used.” (IUCN 2001, 2012).

The Amazon Basin in Brazil represents the largest proportion of rivers (51.2%). Considering the spatial variation in both threats and population genetics, it would probably be useful to include the Amazon Basin seperated into major basins (e.g. HYdroBasin level 4 - Madeira, Negro, etc). But the summaries presented here provide a general overview and enable robust assessment of the results. Additionally, the overall patterns do not change if the Amazon basin is seperated further.

  • The estimates are conservative (best-case) due to the modelling assumptions listed below.
    Endangered (A3bd), is threfore a precautionary IUCN Red List assessment.
    • Important threats are not included (e.g. land use change Zalles et. al. 2021).
    • Losses are potentially buffered by an overly generous 20% ceiling to increases in unaccesible/unhunted populations.
    • Population projections use the earliest likely breeding age. The age at first reproduction was set to 5 years (Norris et. al. 2019).
    • Hunting of adult females was limited to 10% per year in the most pessimistic scenarios.
    • Hunting accessibility was limited to a distance of 48 km from locations with a human population density of 5 or more people per km2. This will underestimate impacts associated with threats such as large scale harvest around urban areas (Chaves et. al. 2021, Tregidgo et. al. 2017) and harvest in less densly populated rural areas (da Silva et al. 2022, Peres 2000).

Methods

To understand the likely future changes in population numbers, spatially explicit population projection scenarios were developed using stage-based Matrix Population Models.
Such scenario based modelling provides a useful and informative global scale assessment for this widespread species, which suffers from a lack of representative and robust evidence on species distribution, abundance and the scale and impact of threats.

The analysis is developed here: https://github.com/darrennorris/TACAR. The workflow, data and associated R code is available here: https://darrennorris.github.io/TACAR/articles/a02_Matrix-population-model-projections.html

The methods used are an extension of Norris et. al. 2019 that includes:

  • Stochastic population projections.
  • Future impacts to populations caused by human acessibility (hunting).

To do

  • Add impacts of actions that reduce river connectivity.

References

Bárcenas‐García A, Michalski F, Gibbs JP, Norris D.2022. Amazonian run‐of‐river dam reservoir impacts underestimated: Evidence from a before–after control–impact study of freshwater turtle nesting areas. Aquatic Conservation: Marine and Freshwater Ecosystems. 32(3):508-22. https://doi.org/10.1002/aqc.3775

Bienvenu F, Legendre S. 2015. A new approach to the generation time in matrix population models. The American Naturalist. 185(6):834-43. https://doi.org/10.1086/681104.

Chaves, W.A., Valle, D., Tavares, A.S., Morcatty, T.Q. and Wilcove, D.S. 2021. Impacts of rural to urban migration, urbanization, and generational change on consumption of wild animals in the Amazon. Conservation Biology, 35: 1186-1197. https://doi.org/10.1111/cobi.13663

da Silva, A. B., et. al. 2022. Patterns of wildlife hunting and trade by local communities in eastern Amazonian floodplains. Ethnobiology and Conservation, 11. https://doi.org/10.15451/ec2022-07-11.16-1-19

IUCN. 2001. IUCN Red List Categories and Criteria: Version 3.1. IUCN Species Survival Commission. IUCN, Gland, Switzerland and Cambridge, U.K.

IUCN. 2012. IUCN Red List Categories and Criteria: Version 3.1. Second edition. IUCN, Gland, Switzerland and Cambridge, UK. Available at www.iucnredlist.org/technical-documents/categories-and-criteria

IUCN. 2024. Guidelines for Using the IUCN Red List Categories and Criteria. Version 16. Prepared by the Standards and Petitions Committee. Downloadable from https://www.iucnredlist.org/documents/RedListGuidelines.pdf.

Jones OR, Barks P, Stott IM, James TD, Levin SC, Petry WK, Capdevila P, Che-Castaldo J, Jackson J, Römer G, Schuette C, Thomas CC, Salguero-Gómez R. 2022. Rcompadre and Rage - two R packages to facilitate the use of the COMPADRE and COMADRE databases and calculation of life history traits from matrix population models. Methods in Ecology and Evolution, 13, 770-781. doi:10.1111/2041-210X.13792.

Norris D, Peres CA, Michalski F, Gibbs JP. 2019. Prospects for freshwater turtle population recovery are catalyzed by pan-Amazonian community-based management. Biological Conservation. 233:51-60. https://doi.org/10.1016/j.biocon.2019.02.022

Stott, I. popdemo vignette: https://cran.r-project.org/web/packages/popdemo/vignettes/popdemo.html

Stott, I., Hodgson, D.J. and Townley, S. 2012. popdemo: an R package for population demography using projection matrix analysis. Methods in Ecology and Evolution, 3: 797-802. https://doi.org/10.1111/j.2041-210X.2012.00222.x

Stubben, C., & Milligan, B. 2007. Estimating and Analyzing Demographic Models Using the popbio Package in R. Journal of Statistical Software, 22(11), 1–23. https://doi.org/10.18637/jss.v022.i11

Tregidgo, D. J., Barlow, J., Pompeu, P. S., de Almeida Rocha, M., & Parry, L. 2017. Rainforest metropolis casts 1,000-km defaunation shadow. Proceedings of the National Academy of Sciences, 114(32), 8655-8659. https://doi.org/10.1073/pnas.161449911

Zalles V, Hansen Matthew C, Potapov Peter V, Parker D, et al. 2021. Rapid expansion of human impact on natural land in South America since 1985. Science Advances, 7: eabg1620. https://doi.org/10.1126/sciadv.abg1620